More recently interviews have become much more structured in format, with the frame of reference very much based closely on the job analysis. All candidates are asked the same predetermined questions with answers scored by the interviewer. Where the questions are situational or job simulating in nature, higher scores represent a more successful the interview, indicating potentially good future job performance.
Harel et al (2003) report increased fairness and standardisation in addition to improved validity and reliability with this method. Interviews which are structured and job-related can be good gages to ascertain cognitive ability and job knowledge, whilst unstructured interviews are more a measure of social skills and personality (Robertson & Smith, 2001).
With the introduction of structured interviews, validity scores are reported to have increased. This is thought due to the preparation required beforehand, where a detailed job analysis is available, forming the basis of the interview questions. When questions asked are on the competencies associated with the job function, it follows that the answers received should give a better indication of ability within that role. It is thought validity could be further improved when interviewers are trained on how to conduct a structured interview, ratings are given for various competencies and not an overall mark, and evaluations made when the interview is completed. In panel situations, each interviewer should complete their evaluation before discussions amongst themselves.
From the perspective of the candidate, Rynes and Connerley (1993) report that interviews are viewed as fair, to the extent that selection procedures are viewed as incomplete when an interview situation is not used. However it is also noted that the preference of the applicant is toward unstructured interviews, possibly because it gives them more freedom in the direction of the interview, and the opportunity to highlight characteristics they feel advantageous.
Aside from interviews, CVs and application forms together comprise the second most frequently used selection technique (Robertson & Smith, 2001). They are often the first communication from a candidate toward a prospective employer, and are essentially self-evaluations. Robertson and Smith (2001) cite several studies (Earl, Bright & Adams, 1998, Bright & Hutton, 2000) which evaluated characteristics of good CVs. All found that the use of very positive competency statements within the CV, covering letter or application form led to an equally positive evaluation of those appraising the application, and an increased probability of invitation to interview or testing situation. Phrases such as ‘highly motivated’, ‘team player’ and ‘high achiever’ are typical of competency statements which can have this affect.
When used independently of other selection techniques, application forms and CVs are poor indicators of future job performance. The one-sided nature of this method lends itself to the low validity observed, as self promoting statements written by the applicant can not be verified employing other techniques.
The use of psychometric tests within the selection process is increasing. Shackleton and Newell (1991) detail an increase in the use of personality tests from 12 to 37% in a five year period, while Doving (2005) reports 51% of employers in Europe are now employing psychological tests as part of their selection procedures. The tests are typically a standardised series of problems or questions, assessing individual characteristics. Each problem can be completed quickly meaning multiple problems can be included for any characteristic, several of which can be assessed at one time. Multiple items provide higher validity then a single indicator of a characteristic.
As with the interview, psychological tests have highest validity when their foundation is rooted in the job analysis. Tests chosen or developed around a specific job function, better predict success within that function than those with less relevance. Logic suggests for a psychological test procedure with reasonable validity, the group difference in test scores should not be significantly different from the observed group difference in job performance.
Psychological tests are less susceptible to interviewer bias, or interviewee impression management than the interview situation. Large numbers of candidates can be assessed at any one time making it an efficient means of processing a large number of applicants. However they have been criticised in the past for lack of relevance and cultural limitations (Compton, Morrissey & Nankervis, 2002). In addition the more rigorous and structured the test (highest validity), the less candidates report liking them.
Psychological tests fall into 3 main groups, interest, personality and ability.
Ability, as measured either by cognitive ability or work sample tests, result in the highest validity results (0.54 and 0.51, respectively) whereas personality and interest rate lower in terms of predicting future job performance, with validity <0.4, (Robertson & Smith, 2001).
Schmidt and Hunter (1984) go so far as to declare that the cognitive ability tests typically used within selection procedures are valid predictors of future performance for “all jobs in all settings”, so long as samples are sufficiently large. It has also been shown that as the complexity of the job increases, the validity of cognitive ability increases. Rynes and Connerley (1993) noted candidates evaluated cognitive ability tests more positively when they appeared job related, and perhaps predictably, when they scored highly.
Personality testing, until relatively recently was not a popular selection technique, nor the source of much research. However the 1990s have seen a substantial increase both in the use of personality assessment with recruitment, and associated research studies (Robertson & Smith, 2001). However opinions differ on the appropriate level of construct best used – should they be maintained as broad measures using a framework such as the big five, or narrow the field, looking at very specific personality factors. Again the key seems to be focusing on specific areas of job performance. Whilst conscientiousness is seen as a valid predictor across all professions, factors such as extraversion are best utilised in very specific career paths.
The use of assessment centres as a selection technique has been widely researched. Doving (2005) found that 29% of European employers use them as part of their recruitment process, whilst in America, in addition to recruitment centres are used both in promotion decisions and also as a tool to enhance job skills (Spector, 2000). The centres comprise a variety of competency based exercises, designed to simulate various job tasks. Assessment typically takes place over several days, with one assessor to two candidates. Generally between 6 and 12 candidates are assessed in one centre on any occasion, and can therefore be both costly and time-consuming. For this reason they are used most often either with a small applicant pool.
The type of exercises used in assessment centres vary widely and range from verbal and analytical reasoning tests, personality questionnaires, role plays and job simulation exercises.
Hough and Oswald (2000, cited in Robertson et al, 2001) noted generally good validity is well established. However several features were noted that could improve the psychometric quality of ratings from assessment centres. These included measuring a few conceptually unique constructs, defining construct based on job, and using assessor with knowledge of the job and training in psychology.
Research also suggests that candidates, on the whole, like assessment centres as a selection tool. Gilliland (1993) proposes that the idea of fairness in enhanced when applicants recognise the relevance of the tests to the job for which they have applied. As a negative, candidate anxiety has been noted as a concern with assessment centres, with some reporting up to 50% candidates’ performance impaired by stress (Teel and DuBois, 1983).
A criticism of both psychological testing and assessment centres focuses around the construct they are measuring. Meta analysis by Scholz and Schuler (1993, cited in Robertson & Smith, 2001) found overall assessment centre ratings to correlate highly with general intelligence, motivation, self confidence and social competence. This result indicates the dominant construct measured in assessment centre tests may be closely related to mental ability. Hunter and Schmidt (2004) in their research on ability tests demonstrate GMA (general mental ability) as a better predictor of future job performance than abilities on individual tasks. Meta-analysis of many validity studies by Hunter and Schmidt and repeatedly highlight cognitive ability as a construct with high validity in terms of predicting future job performance.
Reference checks, together with the application forms and interview, are widely used within the selection process. Although a low statistical validity, 0.26 is presented (Robertson & Smith, 2001), Rynes and Connerley (1993) report it as generally considered fair by candidates. However, in society today, reference checks are essentially used to substantiate an individual’s ability, and to confirm opinions a prospective employer will have formed via other recruitment methods. Stone (2002) considers reference checking a critical undertaking before making a job offer to a candidate.
In conclusion, research into employee selection is extensive, and ongoing. However it has been demonstrated that the primary tool used in recruitment today is the interview, followed by the self-reporting CV or application from. These are tools traditional employed in the practice, and instruments which both the candidate and recruiter are comfortable with. Research has improved the quality and validity as organisations move away from unstructured toward a more structured approach. The use of job analysis as a tool to base the selection procedures upon has been widely adopted. On the whole, employee selection is slowly becoming more scientific, and hopefully moving to becoming more successful, in terms of decisions organisations are making as a result.
References
Caroll, M., Marchington, M., Earnshaw, J., Taylor, S. (1999), "Recruitment in small firms. Processes, methods and problems", Employee Relations, 21, 236-50.
Compton, R.L., Morrissey, W.J. & Nankervis, A.R. (2002). ‘Effective Recruitment and Selection Practices’. CCH Australia Pty Ltd., Australia.
Gilliland, S.W. (1993), The perceived fairness of selection systems: an organizational justice perspective. Academy of Management Review, 18, 4, 694-734.
Harel, G. H., Arditi-Vogel, A & Janz, T. (2003). Comparing the validity and utility of behaviour description interview versus assessment centre ratings. Journal of Psychological managemen, 18, 2, 94-104.
Hunter, J. E. and Schmidt, F. L.(2004). ‘Methods of Meta-Analysis: Correcting Error and Bias in Research Findings’. Sage Publications: Newbury Park.
Nankervis, A.R., Compton, R.L., McCarthy, T.E. 1999, Strategic Human Resource Management, 3rd edn., Nelson Australia.
Newell, S. & Shackleton, V. (1994). The use (and abuse) of psychometric tests in British industry and commerce. Human Resource management Journal, 4, 1.
Rynes, S. L. & Connerley, M. L. (1993). Applicant Reactions to alternative selection procedures. Journal of Business and Psychology, 7, 261-277.
Schmidt, F. L and Hunter, J. E. (1981). Employment Testing; Old Theories and New Research Findings. American Psychologist, 36, 10, 1128-1137.
Spector, P.E. (2000). ‘Industrial & Organisational Psychology, Research and Practice’. Wiley, USA.
Stone, R.J. 2002, Human Resource Management, 4th edn. Wiley, Australia .
Teel, K, DuBois, H (1983), "Participants’ reactions to assessment centers", Personnel Administrator, 28, 85-91.
Webster, E.C, (1964). ‘Decision making in the Employment Interview’. Inductrial Releations Centre, McGill University. Cited by Martin, J. (2005). ‘Organisation behaviour management’. Thomson, London.